subagent

Protocol that evo optimization subagents follow when dispatched from /optimize. Auto-loaded by spawned subagents via their host's skill loader. The orchestrator may also invoke this skill to understand the brief shape its dispatched subagents expect + what they're required to emit -- useful when writing briefs or debugging a subagent's behavior.

Quality

85%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Low

Low-risk findings worth noting

Quality

Content

85%Weight 40%Scale 1-3

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

The protocol is exceptionally actionable with a clear, checkpointed workflow and well-signaled references. The main cost is conciseness: the inline detail is large and could be trimmed or offloaded to reference files.

Suggestions

Move the commit-strategy (--i-staged-new-files) and remote-recovery internals into references/cli-quick-reference.md, keeping only the decision rule inline to cut ~50 lines.

Condense the telemetry feedback example block to the trigger conditions and a one-line command skeleton rather than a full flag-by-flag invocation.

Dimension	Reasoning	Score
Conciseness	The body is highly operational and mostly earns its tokens, but at ~400 lines it carries notable explanatory padding (commit-strategy variants, recovery internals, telemetry rationale) that could be tightened or pushed to references.	2 / 3
Actionability	Provides copy-paste-ready commands ('evo new --parent <id> -m "<hypothesis>"', 'evo run <exp_id>') and exact Task tool invocation templates, leaving little abstract.	3 / 3
Workflow Clarity	The Iteration Loop is a numbered 0–9 sequence with explicit validation checkpoints (pre/post verifier, gate checks), bounded retry feedback loops, and outcome classification guiding error recovery.	3 / 3
Progressive Disclosure	Cross-skill references are well-signaled one-level-deep with per-line triggering conditions ('sdk_python.py — wiring per-task instrumentation', 'evo-wait.md — any time you need to wait'); sections are clearly organized for navigation.	3 / 3
	Total	11 / 12 Passed

Description

85%Weight 40%Scale 1-3

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

The description clearly defines a distinct niche and answers both what and when, with concrete actions and third-person voice. Its weakness is trigger-term quality: heavy domain jargon over natural user phrasing.

Suggestions

Add plain-language trigger terms alongside the jargon (e.g., 'Use when tuning or optimizing an evo experiment, dispatching optimization subagents, or debugging subagent behavior') so the skill surfaces from natural requests.

Lead with a concise capability sentence before the auto-loading mechanics, so the core action is readable without parsing 'host's skill loader' internals.

Dimension	Reasoning	Score
Specificity	Lists multiple concrete actions — 'subagents follow when dispatched from /optimize', 'Auto-loaded by spawned subagents', 'invoke this skill to understand the brief shape', 'debugging a subagent's behavior' — rather than vague language.	3 / 3
Completeness	Explicitly answers what it does ('Protocol that evo optimization subagents follow') and when to use it ('when dispatched from /optimize', 'useful when writing briefs or debugging'), satisfying both the what and when requirements.	3 / 3
Trigger Term Quality	Contains some relevant keywords ('/optimize', 'subagent', 'brief', 'orchestrator') but is dominated by technical jargon ('host's skill loader', 'evo optimization subagents') that users would not naturally say; coverage of natural variations is thin.	2 / 3
Distinctiveness Conflict Risk	The niche is sharply scoped to 'evo optimization subagents' dispatched from /optimize, with distinct triggers unlikely to overlap with unrelated skills.	3 / 3
	Total	11 / 12 Passed

Validation

87%

Warnings & errors only

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 14 / 16 Passed

Validation for skill structure

Criteria	Description	Result
frontmatter_unknown_keys	Unknown frontmatter key(s) found; consider removing or moving to metadata	Warning
referenced_paths_exist	Referenced path issues: 2 missing	Warning

	Total	14 / 16 Passed

Repository: evo-hq/evo
Path: plugins/evo/npm/skills/subagent/SKILL.md
Commit: ab5fbd6

Reviewed: 12 days ago

Table of Contents

Discovery Implementation Validation

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.